Automated summary
Diagnoses failing GitHub Actions checks on a pull request or branch by listing failures, ranking falsifiable hypotheses, confirming the root cause with log evidence, then applying and verifying the smallest authorized fix.
Strengths
- Requires hypotheses and the fastest falsification command to be stated before any investigation begins.
- Demands log evidence before a root cause or fix is proposed.
- Constrains changes to the smallest authorized edit and requires the exact CI verification command from the workflow YAML.
- Clearly separates fix requests from review-only requests that should not edit files.
Limitations
- Covers only GitHub Actions and relies on `gh` CLI authentication and permissions.
- Does not describe flaky-failure detection, job reruns, or infrastructure remediation.
- Allowed commands are limited to `gh pr`, `gh run`, `git branch`, and `git log`, which may restrict some investigations.
Security analysis
SafeThe skill is a read-only CI diagnosis workflow built entirely on benign GitHub CLI and git inspection commands; there is no rm, curl|sh, credential handling, obfuscation, or instruction-override content. The only notable concerns are the mildly permissive authorization phrasing and the Step 6 suggestion to run the project's CI command, which exceeds the declared tool allowlist but is a normal, non-destructive developer action.
- •Step 5 contains authorization-assumption language ('existing authorization counts', 'Ask only at the root AGENTS.md boundaries') that could nudge an agent to make edits without explicit confirmation; impact is limited because the declared tool surface is read-only.
- •Step 6 asks the agent to run 'the same command CI runs' (e.g. the workflow's build/test command), which falls outside the declared `gh`/`git` read-only tool allowlist — a scope expansion, though a conventional and non-destructive one.
- •No destructive, exfiltrating, or obfuscated actions: all declared commands (`gh pr checks`, `gh pr view`, `gh run list/view --log-failed`, `git branch --show-current`, `git log`) are read-only inspection; logs may contain environment values but nothing instructs sending them anywhere.
Examples
CI is red on PR #482. List the failing checks, rank your hypotheses, and confirm the root cause before changing anything.The lint job on my branch failed. Run CI triage and apply the smallest authorized fix, then verify it with the exact command CI uses.Why did the test workflow fail on feature/login? Investigate the logs, report the confirmed cause, and propose a fix without editing files.name: ci-triage description: "Diagnoses failing GitHub Actions checks on a PR or branch: lists the failures, ranks falsifiable hypotheses, confirms the root cause with log evidence, then applies and verifies the smallest authorized fix. Use when CI is red, a check or workflow run fails, or the user asks why a build, lint or test job failed." allowed-tools: Bash(gh pr *), Bash(gh run *), Bash(git branch *), Bash(git log *)
CI Triage
Step 1 — List failures
Run gh pr checks on the current PR and list every failing check by name. For a push-triggered run with no PR, use gh run list --branch <branch> --limit 5 instead.
Step 2 — Hypotheses first, before any investigation
For each failing check, enumerate your top 3 hypotheses with the single fastest command to falsify each (usually starting with gh run view <id> --log-failed). Present this list to the user. Do not start running commands or reading files yet.
Step 3 — Investigate Test each hypothesis from cheapest to most expensive; read-only investigation needs no additional confirmation. Stop and report findings after each hypothesis before moving to the next.
Step 4 — Confirm root cause State the confirmed root cause clearly. Do not guess or propose a fix until you have evidence.
Step 5 — Apply the authorized fix For a fix request, make the smallest supported change within the task. For a review-only request, report the proposed fix without editing. Ask only at the root AGENTS.md boundaries; existing authorization counts.
Step 6 — Verification command After the fix is applied, run and report the exact command that confirms it resolves the failure before pushing — the same command CI runs (from the workflow YAML), not an approximation.
Current PR info (Claude Code fills these in; in Codex, run the commands):
- Branch: !
git branch --show-current - Recent commits: !
git log --oneline -5
From the file to a first result
The download contains exactly the content displayed on this page. This copy may differ from the latest version in the source repository.
View the source and supporting filesThis download contains SKILL.md only. Any scripts, references, templates and dependencies are not included. Get the required files from the source, preserving their directory structure, before continuing.
1. Download the file
Download SKILL.mdFrom the root of a test project, save the file as .skill-downloads/ci-failure-triage-mvg6txae/SKILL.md, or use the command below. This folder is for inspecting the skill before installation.
Command for macOS/Linux or WSL: requires a POSIX shell and curl. On Windows without WSL, use the download button and create the folder manually.
mkdir -p '.skill-downloads/ci-failure-triage-mvg6txae' && curl --fail --show-error --location 'https://skills-guides.com/api/skills/ci-failure-triage-mvg6txae/download' --output '.skill-downloads/ci-failure-triage-mvg6txae/SKILL.md.part' && mv '.skill-downloads/ci-failure-triage-mvg6txae/SKILL.md.part' '.skill-downloads/ci-failure-triage-mvg6txae/SKILL.md'2. Check prerequisites
Read SKILL.md before using it. If it references scripts or external services, inspect them and prepare the required access in your test environment.
Tools declared by the author (check availability in your agent): Bash(gh pr *), Bash(gh run *), Bash(git branch *), Bash(git log *)
No recognized prerequisites section in this file. This does not mean it is self-contained: libraries, tools and access requirements may be mentioned elsewhere in the content or source repository.
The platforms on this page are catalog indications. The paths below follow tool documentation; they do not prove this skill has been run in those tools. No minimum tool version is confirmed for this listing.
3. Load the skill
After inspection, copy the complete folder into your project using one of the commands below, then open a new tool session. The command stops if a folder with the same name already exists: compare the files before replacing it.
Claude Code
test -f '.skill-downloads/ci-failure-triage-mvg6txae/SKILL.md' && test -s '.skill-downloads/ci-failure-triage-mvg6txae/SKILL.md' && mkdir -p '.claude/skills' && test ! -e '.claude/skills/ci-triage' && cp -R '.skill-downloads/ci-failure-triage-mvg6txae' '.claude/skills/ci-triage'Codex
test -f '.skill-downloads/ci-failure-triage-mvg6txae/SKILL.md' && test -s '.skill-downloads/ci-failure-triage-mvg6txae/SKILL.md' && mkdir -p '.agents/skills' && test ! -e '.agents/skills/ci-triage' && cp -R '.skill-downloads/ci-failure-triage-mvg6txae' '.agents/skills/ci-triage'OpenSkills
OpenSkills must already be installed. The local install/read procedure is checked with OpenSkills 1.5.0, which requires Node.js ≥ 20.6.0. Catalog slugs are not resolved automatically.
Official OpenSkills installationopenskills install './.skill-downloads/ci-failure-triage-mvg6txae'openskills read 'ci-failure-triage-mvg6txae'Reading should display the installed file content and its base directory. It does not run the skill or install its dependencies.
For an agent that reads AGENTS.md, then sync the index. This command changes AGENTS.md: review the diff before using it.
openskills syncManual method or another tool: attach SKILL.md to a conversation, or open it in an agent that can read files. Include any required supporting resources. No universal folder is recognized by every tool.
4. Try a first task
Use a short, synthetic input: an API endpoint, a small document or an interface brief. In a new conversation, attach the file or give the agent its path, then adapt this request:
Read .skill-downloads/ci-failure-triage-mvg6txae/SKILL.md and apply its instructions to this example: [describe a test input]. Identify any missing files, dependencies or access. Produce the result requested by the skill without inventing data or running unreviewed scripts.Expected result: output matching the skill instructions, or a precise list of missing prerequisites. A read confirmation alone does not validate a first use. If an external resource is unavailable, that path remains untested.
Configure a Chat Channel with SecretRefs
DevOps
Set up chat channels using non-interactive commands, with secrets only as SecretRefs. Covers gather, mutate, repair, and prove steps for reliable delivery.
Investigate ClickHouse CI Failures
DevOps
Investigate a ClickHouse CI failure end-to-end from a PR or S3 report URL. Fetches failed tests, classifies flaky vs real regression, and checks for existing issues/fixes.
Ansible PR Review
DevOps
Reviews an Ansible PR following the project's standardized process, including changelog checks, test coverage validation, and feedback tracking.